Skilled trades training in the age of AI: job and openings data

Skilled trades training in the age of AI: what the research can and cannot show

Artificial intelligence is changing how many kinds of work are discussed, but the available research does not support a simple prediction that AI will eliminate skilled trades jobs. For students, career changers, parents, and counselors considering skilled trades training in the age of AI, the more useful question is narrower: Which tasks might technology handle, and which still depend on human judgment, safety decisions, and hands-on work?

Research from MIT Sloan suggests that AI is more likely to complement human workers than replace them. The study examines tasks across occupations, not electricians, mechanics, or production workers specifically, so it should be used as a framework for thinking about training, not as proof of future hiring in any one trade.

What the MIT research says about AI and skilled trades jobs

The MIT research distinguishes between automation and augmentation. Automation moves a task from a person to a machine. Augmentation uses a machine to increase a worker’s productivity or make work possible that could not be done before, according to MIT Sloan.

That difference matters in skilled work. A technology tool might help identify a possible equipment problem, organize information, or flag an unusual reading. A worker may still need to inspect the physical system, decide whether the information is reliable, follow safety procedures, and choose what to do next. This is a hypothetical example, not evidence that employers are currently using a particular AI system.

MIT Sloan’s researchers, Roberto Rigobon and Isabella Loaiza, studied the structure of tasks across a variety of occupations. Their analysis used an EPOCH index, a risk-of-substitution score, and a potential-for-augmentation score to examine how human capabilities and technology may interact, according to MIT Sloan.

The research found more human-intensive tasks, and more frequent performance of those tasks, between 2016 and 2024. It also found that tasks added to the O*NET data set in 2024 had higher levels of EPOCH capabilities than tasks that existed before then or disappeared in 2024, according to MIT Sloan.

That finding does not predict what will happen in a particular apprenticeship or workplace. It does suggest that training focused only on repeating technical steps may be incomplete as tools become more capable.

The human capabilities behind the EPOCH framework

EPOCH describes five groups of capabilities that machines may handle less effectively:

• Empathy and emotional intelligence
• Presence, networking, and connectedness
• Opinion, judgment, and ethics
• Creativity and imagination
• Hope, vision, and leadership

These categories do not mean that technical skills are unimportant. A skilled worker still needs to understand tools, materials, procedures, and safety rules. The point is that technical knowledge and human decision-making can work together.

A worker may need to communicate with a customer, coordinate with another crew, recognize when a situation does not match the usual pattern, or decide whether a repair is safe to complete. Those decisions can involve incomplete information and consequences that are difficult to reduce to a routine instruction.

MIT’s research examined the statistical limitations of AI tools as well. The researchers note that such tools can perform poorly when data are biased or limited, when a task requires extrapolating far beyond the training data, or when a moral dilemma is involved, according to MIT Sloan.

For a student, the practical lesson is not to avoid technology. It is to learn how to question an output, compare it with physical evidence, and recognize when a person must make the final decision.

Why the connection to trade training remains limited

The MIT study is economy-wide. It does not isolate electricians, industrial maintenance workers, welders, or other specific skilled trades. It also does not establish that AI is creating demand for trade workers or changing apprenticeship requirements.

That distinction should remain visible when comparing training programs. A broad finding about human-machine collaboration cannot answer every local question, such as whether a particular employer uses digital diagnostic equipment or whether a state licensing board recognizes a specific course.

The research does offer a way to evaluate what a program teaches. A strong pathway may need to develop both occupation-specific skills and the ability to apply those skills when conditions change. That can include reading technical information, identifying an error, communicating clearly, following safety rules, and making careful decisions when an automated recommendation is incomplete.

These are not separate from hands-on training. They are part of using hands-on training responsibly.

What skilled trades training should help students examine

The research does not show whether technical schools or apprenticeship sponsors are adding AI literacy to their curricula. It does not show whether employers are changing job descriptions or adopting particular diagnostic systems. Those questions require information from individual training providers, employers, apprenticeship sponsors, or trade associations, none of which is included in the supplied research.

Readers can still ask useful, specific questions before enrolling:

• Does the program prepare students for the exact occupation under consideration?
• Which hands-on tasks will students practice, and how will instructors assess them?
• Does the curriculum include safety procedures, troubleshooting, technical documentation, and supervised practice?
• Are digital tools or diagnostic systems part of the training, and if so, how are students taught to verify their results?
• What credential, exam, or license does the program prepare students to pursue?
• Which requirements come from the state, which come from the program, and which come from an employer?
• Is the schedule realistic for a student who works, provides care, or cannot attend full time?

Requirements vary by occupation, state, school, and employer. A program’s answer should be checked against the relevant official licensing or credentialing office before enrollment.

How to think about the future of skilled trades careers

The available evidence supports a careful conclusion. MIT Sloan’s research points toward AI complementing workers when people contribute capabilities that technology may not handle well, including judgment, ethics, creativity, communication, and leadership. It does not prove that every skilled trade will expand, that every program will lead to employment, or that automation will have no effect.

For students and career changers, that uncertainty is a reason to compare pathways closely, not a reason to treat trade work as either automatically safe or automatically threatened. Look for training that combines core technical instruction with troubleshooting, safety, communication, and responsible use of digital tools.

Before making a decision, ask the program for its current curriculum, credential expectations, hands-on requirements, schedule, and licensing information. Then verify those details with the appropriate state or program authority. That process will provide a more useful starting point than either claim that AI will replace skilled workers or the promise that technology will guarantee new opportunities.